A

SEMINAR REPORT
ON
SECURE ATM BY IMAGE PROCESSING

SUMITTED IN PATIAL FUFILMENT OF THE

REQUIREMET FOR THE AWARD OF NATIONAL
DIPLOMA.
FEDERAL POLYTECHNIC ADO-EKITI

SUBMITTED BY
POPOOLA JULIUS KAYODE
FPA/CS/20/2-0212
 ACKNOWLEDGEMENT
I would like to thank my coordinator for giving me the opportunity to broaden my

knowledge and leading me along the way, it has helped me to know more.

Secondly, I would like to thank my parents who helped me through all my work

and prayed for me along the way and helping me to eliminate all problems I may

face.

I would like to thank my friends and all students of computer science for their

cooperation with me.

Finally, I would like to thank GOD ALMIGHTY for guiding and protecting me

along the way and through my journey.

 CONTENT

1 ABSTRACT

2 INTRODUCTION

3 TECNIQUE OF IMAGE PROCESSING

4 METHODOLOGY

5 DRAWBACKS OF BIOMETRICS

6 ADVANTAGES AND DISADVANTAGES

7 CONCLUSION

8 REFRENCES
ABSTRACT

Every biometric system has its limitations. Therefore, identification based on multiple

biometrics is an emerging trend as multi model biometrics can help to provide a more

balanced and effective solution to the systems which involves the use of more than one

biometric system.

My contribution to the above subject is that an algorithm can be developed for the bank

security. This security measure is assured by using finger scan, voice scan, hand scan,

geometry scan and also with the request of the password given by the bank for a

particular used when necessary.

This biometric technology allows determination and verification of one’s identity with

the aid of physical characteristics, in simpler words, the human body becomes his/her

password. Algorithms have been proposed by taking biometric techniques to authenticate

an ATM account holder, which will enable a Secure ATM by Image Processing. This

biometric method is now being used in many places both private and public.
INTRODUCTION

Programmed AUTOMATED TELLER MACHINE (ATM) is advanced Information

innovation gadget. That gadget gives the clients and clients of monetary association with

go to the budgetary exchanges in open space for clients without the need human agent. In

ATMs, Customer is recognized by utilizing the ATM Card and that card embedded with

chip, that atm card contain an exceptional card number and some other security data

about the client.

All biometric has its every own impediments. That is the spot dependent on numerous

biometrics is seem pattern as multi model concede an additional arrangements and they

are adjusted as well. A calculation on banking security utilizing bioscience for its security

reason. The security support by utilizing finger filter, hand check and by mentioning for

the secret key given by the bank for a that client when its important. Biometrics

innovation permits persistence and demonstrate of once personality through physical

qualities. It transforms your body into your secret key. We talked about different

biometric procedures. That is all are retina filter, finger examine, facial output, hand

check and so on there are two calculations have been configuration by taking biometric

systems to verification an ATM account holder or the record client, empowering a safe
ATM by picture handling. Biometrics is currently accessible in any resembled in

different open and private segments moreover.

No more issues if passwords and I'd codes have been overlooked, biometrics is the

innovation that deals with it, making your body your secret key. Typically, to make your

mystery word assurance and advancement controls logically exhaustive, the more

problematic it will be for customers to review their passwords. Unfortunately, to stop

essential software engineer strikes on the framework, serious mystery key rules are

required. The current accessible age security issue is viewed as the fundamental TCP/IP

encryptions and different variables that are given by the utilizing system. Be that as it

may, there was a considerable lot of predictable distinguishing proof of each one

separately, at that point the recently created innovation is Biometrics. Biometrics can be

utilized to evade the unapproved access to ATM, PDAs, home security frameworks,

entryway locks, keen cards, work area PC's workstations. This paper bound the data

respecting the 'picture preparing'. What's more, there is one of the significant use of

picture handling is the 'biometrics'.

LITERATURE REVIEW OF IMAGE PROCESSING

TECHNIQUES

The utilization is that no more issues if client overlooked there passwords and id codes

additionally, in truth, biometrics is the innovation that deals with that which transforms

your body into your passwords. The all the more requesting you make your own secret

key decision and development the guidelines done the more trouble to the clients can

have in recalling their passwords. However, severe and hard secret word rules are

important to keep away from simple programmer assaults on the system.

The essential downside with secret key is two folds. And furthermore, they are

assignable, that can be recorded onto the paper and that can exchanged to somebody who

shouldn't have them. Furthermore, they will overlooked. As of late examination proposes

that the overlooked secret phrase will prize as much as US$ 340 for each occasion, the

possibility and expenses of course of action passwords are a central point. In the reality

the significant and principle requirement for additional dimension of security has offered

ascend to field of

"BIOMETRICS"
ATM authentication system based on fingerprint. We can refer to fingerprint scanning

technology with the help of this paper. Biometric technology used in a wide range of

physical access and logical access applications is most commonly used in finger scanning

technology. The characteristics and patterns of all fingerprints are unique. A normal

fingerprint consists of spaces, lines. These lines are called ridges, while valleys are called

the spaces between the ridges.

This provides a summary of the techniques of recognition of the human or user face. And

that applies a lot to the front faces, there are advantages and disadvantages of each

method are also given. The methods considered are individual features and semantic

networks, as well as the dynamic link architecture used to verify the face of the person,

hidden Markov model, matching computational features, and matching disorder. These

approaches are analyzed in terms of the image processing facial representations they

used.
 METHODOLOGY

The two classes of biometric strategies are:- Physiology based procedures that measure

the physiological qualities of an individual. These incorporate unique mark confirmation,

iris examination, facial investigation, hand geometry vein designs, ear recognition, smell

discovery, and DNA design examination. The social procedures which measure the

conduct of an individual. These incorporate written by hand signature check and

discourse examination.

The entire method of Image processing and starting from the receiving of visual

data to the giving out of description of the scene, can be divided into three major

stages which are represented as major sub areas. They are:

1. Discretization and illustration: changing visual data into a discrete form: suitable for

computer processing: approximating visual data to avoid wasting storage space as well as

time requirement in consequent processing.

2. Processing: improving image quality by filtering etc; compressing information to avoid

wasting storage and channel capacity throughout transmission.

3. Analysis: Extracting image features; qualifying shapes, recognition and registration too.
No more issues if passwords and Id codes have been overlooked, biometrics is the

innovation that deals with it, making your body your secret word. Typically. To make

your secret key determination and development runs progressively thorough, the more

troublesome it will be for clients to recall their passwords. Sadly, to stop basic

programmer assaults on the system, severe secret key standards are required.

Fig. Below shows the types of biometrics

Fig 1). Types of Biometrics

1) FINGER PRINT SCAN

Biometric finger scanning is based on the characteristic features of the human fingerprint.

In forensic applications, fingerprints are used: large-scale, one-to many searches of up to

millions of fingerprints on database. Most widely used biometrics technology is

fingerprint recognition.
First step is called fingerprint template formation when the original of your fingerprint

extracted from optical sensor. The fingerprint is form in an 8 bit grace skills. The 8-bit

skills converted into 0 and 1 depending on the nearby pixels. this is done with the

algorithm After getting the 8-bit skills convert the original fingerprint lines into the

thinning lines after that the bifurcation is marked.15 to 20 images are marked. Then the

range of fingerprint template is of 240 to 320 pixel. Each template is in the range of 200

to 500 bytes.

The next step is matching the finger prints with the database with the current finger print.

This complete process is handled by the Digital signal processing at 400 mhz. this all

process done in few sec. There are two types of fingerprint scanning.
Optical scanning

System can take the scan of user palm then this print can register in the database when

user can use this technique again then scan the hand then match with database scan. If

they are match then user can use the system this type is not much secure any other fraud

user can use this fraud user can take users fingerprint easily and use it.

Capacitor scanning

Capacitor sensor use in this scanning breadth if this sensor is 1mm(micro meter) or less

than 1mm.using this feature the capacitor can take print of users fingerprint.in that the use

scan only limited points. When this process is start then some capacitor are activate and

some are not activated. Then take and print of users fingerprint. Capacitor can search

only 45 point and catch it.

2) IRIS SCAN

The iris has colored streaks and lines radiating from the eye's pupil. After DNA, the iris

provides the most complete biometric data. The iris has information that is more unique

than any other organ in the body.

The iris scan is safe than the fingerprint ordinary camera can take a picture of users iris.

Use users iris picture can use for one code of that user and that code user can use after for

using the system. In the iris scan capture a 240 points are scan.

The iris is the hued ring of muscle that opens and closes the student of the eye like a

camera screen. The shaded example of our irises is resolved hereditarily when we're in

the belly however not full fledged until we're matured around two. It originates from a

color called melanin— more melanin gives you browner eyes and less creates bluer eyes.

In spite of the fact that we talk about individuals having "blue eyes," "green eyes,"

"darker eyes," or whatever, in all actuality the shading and example of individuals' eyes is

incredibly mind boggling and totally interesting: the examples of one individual's two

eyes are very unique in relation to one another and even hereditarily indistinguishable

twins have distinctive iris designs. In the first place, every one of the general population

the framework has to think about must have their eyes examined. This irregular

procedure is called enlistment. Every individual stands before a camera and has their eyes

carefully shot with both standard light and undetectable infrared (a kind of light utilized

in night vision frameworks that has a somewhat longer wavelength than customary red

light). In iris acknowledgment, infrared shows up the one of a kind highlights of dimly
shaded eyes that don't emerge obviously in standard light. These two advanced photos are

then broke down by a PC that expels superfluous subtleties, (for example, eyelashes) and

recognizes around 240 extraordinary highlights (around multiple times more "purposes of

examination" as unique mark frameworks use). These highlights, remarkable to each eye,

are transformed into a basic, 512-digit number considered an Iris Code® that is put away,

close by your name and different subtleties, in a PC database. The enlistment procedure is

totally programmed and as a rule takes close to two or three minutes.

When you're put away in the framework, it's a basic issue to check your recognize. You

just remain before another iris scanner and have your eye shot once more. The framework

rapidly forms the picture and concentrates your Iris Code®, before looking at it against

the hundreds thousands, or millions put away in its database. On the off chance that your

code matches one of the put away ones, you're decidedly distinguished; if not,

unfortunate news! It either implies you're not known to the framework or you're not

whom you guarantee to be.

 Fig 3). Area of focus in the eye.

3) FACE SCAN

Like all biometrics arrangements, face acknowledgment innovation measures and

matches the interesting attributes for the reasons for distinguishing proof or validation.

Regularly utilizing an advanced or associated camera, facial acknowledgment

programming can identify faces in pictures, evaluate their highlights, and afterward

coordinate them against put away layouts in a database. Face examining biometric tech is

unbelievably adaptable and this is reflected in its wide scope of potential applications.

Face biometrics can possibly be incorporated anyplace you can locate an advanced

camera. Law requirement offices the world over use biometric programming to examine

faces in CCTV film, just as to recognize people of enthusiasm for the field. Outskirt

control arrangements use face acknowledgment to check the personalities of explorers. It

even has purchaser applications. We are additionally observing face biometrics in the

advanced world, with Facebook, Shutter Stock, and other social stages that look to sort
out amazing measures of rich picture information by distinguishing the general

population caught in them. Facial acknowledgment doesn't simply manage hard

characters, yet in addition can assemble statistic information on groups. This has made

face biometrics arrangements much looked for after in the retail showcasing industry.

Facial structure is likewise a physiological methodology that can be utilized for

individual recognizable proof and validation. Human facial structure is an individual

trademark. Facial acknowledgment biometrics utilizes this reality to recognize and verify

people. Human minds have common capacity to recollect and recognize various

countenances. We distinguish and confirm individuals just by perceiving their face once a

day. We perceive our family, companions, partners, and pets principally by their facial

structure. Facial acknowledgment framework can recognize individuals by handling their

advanced pictures if their facial acknowledgment personality has been pre-built up. The

framework exploits advanced pictures or still casings from a video source, which are

taken through the facial acknowledgment calculation. This calculation extricates

information out of facial qualities like position and state of eyes, nose, cheekbones and

jaw. It can likewise gauge remove between these qualities and mapped information is put

away in a database. This framework can be helpful in distinguishing individuals in group

like air terminal terminals, railroad stations, and so forth. Facial acknowledgment

frameworks can catch various pictures in a second, contrast them and what is put away in

the database and delivered results.

 Types Accuracy Ease Use Acceptance

Fingerprint High Medium Low

Face Low High High

Iris High Medium Medium

Table. Comparison of biometrics types.

DRAWBACKS ON THE USE OF BIOMETRICS

1. A major problem with biometrics is how and where to store the user’s template. Its

storage introduces privacy concerns. On the other hand, storing the template on a

smart card enhances individual privacy and increases protection from attack,

because individual users control their own templates.

 2. Inside the eye is a very delicate area and so, many people are very hesitating to use

laser eye device.

3. A major problem with biometrics is how and where to store the user’s template. Its

storage introduces privacy concerns. On the other hand, storing the template on a

smart card enhances individual privacy and increases protection from attack,

because individual users control their own templates.

4. Time Condition. - Biometrics may change as an individual ages as the features of

their face which have been recorded might change.

5. Environment:- Environmental conditions may either after the biometric directly

(for example, if a finger is cut and scarred) or interfere with the data collection (for

instance, background noise when using a voice biometric).

ADVANTAGES OF BIOMETRIC AUTHENTICATION

Biometric authentication and its uses in modern-day tech and digital applications has a

number of advantages:
 1. High security and assurance – Biometric identification provides the answers

to “something a person has and is” and helps verify identity

2. User Experience – Convenient and fast

3. Non-transferrable – Everyone has access to a unique set of biometrics

4. Spoof-proof – Biometrics are hard to fake or steal

DISADVANTAGES OF BIOMETRIC AUTHENTICATION

Despite increased security, efficiency, and convenience, biometric authentication and its

uses in modern-day tech and digital applications also has disadvantages:

1. Costs – Significant investment needed in biometrics for security

2. Data breaches – Biometric databases can still be hacked

3. Tracking and data – Biometric devices like facial recognition systems can

limit privacy for users

4. Bias – Machine learning and algorithms must be very advanced to minimize

biometric demographic bias

5. False positives and inaccuracy – False rejects and false accepts can still occur

preventing select users from accessing systems

FUTURE OF BIOMETRICS

Biometric technology is one area that no segment of the IT industry can afford to ignore,

biometrics provides security benefits across the spectrum, from IT vendors to end users,

and from security system developers to security system users. All these industry sectors

must evaluate the costs and benefits of implementing such security measures. :

Biometric technology is one area that no segment of the IT industry can afford to ignore,

biometrics provides security benefits across the spectrum, from IT vendors to end users,

and from security system developers to security system users. All these industry sectors

must evaluate the costs and benefits of implementing such security measures.

CONCLUSION:

We have developed a rapid fingerprint enhancement algorithm that can adaptively

improve the clarification of ridge and furrow structures based on the estimated local ridge

orientation and ridge frequency from the images input. Bank uses fingerprint readers for

ATM authorization and becomes more common in grocery stores where they are used to

automatically recognize a registry customer and bill their credit card or debit account
Enhancement algorithm using the minutiae's goodness index and input fingerprint

verification accuracy. The Enhancement algorithm is an input fingerprint verification

technique. The algorithm also identifies, and removes from further processing, the

unrecoverable corrupted regions in the fingerprint.

Multimodal biometrics and two-tier safety ensure a higher level of safety. Error rates

such as False Acceptance Rate and False Reject Rate have been reduced, preventing

different types of attacks in the ATM system and reducing fraudulent activity. It is

strictly avoided that hackers have the opportunity to use fake biometrics to act as an

authorized user, which makes the ATM system more secure. But compared to the

existing ATM system, the cost of designing and implementing this type of system. The

proposed ATM's security algorithm implies a secure biometric world, enabling image

processing to secure ATM'S. Biometrics technologies are still emerging. This article

presents a snapshot of the dynamics currently underway in this emerging biometric

market and we hope that it will help all possible alternatives in the acquisition of new

biometric technologies
REFRENCES

Biometric Devices ||Star Link

Fingerprint Based Biometric ATM Authentication System

https://www.ijraset.com
https://scialert.net/fulltext

www.google.com

www.wikipedia.com

https://www.researchgate.net/figure

Oxford University Press

Cambridge research institute (Dr Park Jin)